Journal of Chromatography B, 1039 (2016) 66–73

Contents lists available at ScienceDirect

Journal of Chromatography B
journal homepage: www.elsevier.com/locate/chromb

Use of response surface methodology for partitioning, one-step

purification of alkaline extracellular lipase from Penicillium candidum
(PCA 1/TT031)
Amaal M. Alhelli a,b , Mohd Yazid Abdul Manap a,c , Abdulkarim Sabo Mohammed a ,
Hamed Mirhosseini a , Eilaf Suliman a , Zahra Shad a , Nameer Khairullah Mohammed a ,
Anis Shobirin Meor Hussin a,c,∗
a
Faculty of Food Science and Technology, University Putra Malaysia, 43400 UPM Serdang, Selangor Darul Ehsan, Malaysia
b
Institute of Technology, Middle Technical University, 29008 Alzafaranya, Baghdad, Iraq
c
Halal Products Research Institute, University Putra Malaysia, 43400 UPM Serdang, Selangor Darul Ehsan, Malaysia

a r t i c l e i n f o a b s t r a c t

Article history: This report shows the partitioning and purification of alkaline extracellular lipase from Penicillium can-
Received 29 June 2016 didum (PCA 1/TT031) by solid-state fermentation (SSF). In the present analysis, some of the important
Received in revised form parameters such as PEG concentration, PEG molecular mass, salt concentration and buffer concentration
27 September 2016
were optimised through the response surface methodology (RSM). The optimum aqueous two-phase
Accepted 26 October 2016
systems (ATPS) environment consisted of 13.8% (w/w) phosphate buffer, 9.2% (w/w) PEG-3000 and 3.3%
Available online 5 November 2016
(w/w) NaCl at 25 ◦ C. The RSM approach was proved to be the most suitable methodology for the recovery
of desired enzymes. In this method, the enzyme partitioned into the top phase of the PEG-buffer–NaCl
Keywords:
Aqueous two phase systems
ATPS. Under this experimental environment, the purification factor was found to be 33.9, the parti-
Penicillium candidum tion coefficient was 4.0 and the yield was found to be 84.0% of lipase. Moreover, the experimental and
RSM predicted results were in considerable agreement, which established the reliability and validity of the
Lipase proposed model. The ATPS methodology is proven to be effective for the primary recovery of lipase at a
Solid state fermentation low cost with a large loading capacity and possibility of linear scale up. In addition to using the existing
methodologies for improving enzyme production, the use of statistical optimisation of the constituents
of phases through RSM continues to be the basic and practical method.
© 2016 Elsevier B.V. All rights reserved.

1. Introduction named Penicillium candidum [7]. Solid-state fermentation utilises

According to biotechnology, lipases triacylglycerol hydrolases
(E.C. 3.1.1.3) are considered as an extremely vital and predomi-
nantly multipurpose group of enzymes, since they can function
as a catalyst in both aqueous and non-aqueous media for a wide
range of reactions. In general, lipases are secreted by yeasts [1],
bacteria [2], and fungi [3]. These lipases are then widely used in
different industries such as leather, dairy, food, and pharmaceuti-
cals [4–6]. An inducible, extracellular lipase that is extensively used
in the dairy industry for flavouring is secreted by a deuteromycete
∗ Corresponding author at: Department of Food Technology, Faculty of Food Sci-
ence and Technology, University Putra Malaysia, 43400 UPM Serdang, Selangor,
Malaysia.
E-mail address: shobirin@upm.edu.my (A.S. Meor Hussin).
the natural agro-industrial residues as a substrate and hence is con-
sidered as a favourable process for secretion of lipase [8]. Today,
aqueous two- phase systems (ATPSs) have been established as
one of the topmost economical downstream processing techniques
for the retrieval of biomolecules such as nucleic acids, proteins,
amino acids, enzymes and microorganisms, from the impurities
and contaminants [9]. ATPS is a very popular technique as it offers
a number of advantages over conventional purification procedures
used for the production of industrial enzymes such as a minor
operational environment, ease of scaling up, use of cost-effective
materials, high speed of processing, limited denaturation of the
proteins and low interfacial tension [10] between two immisci-
ble phases [11]. On mixing aqueous solutions of two hydrophilic
polymers, or a salt and a polymer, above a specific critical con-
centration, ATPS immediately initiates the processing [12]. Since
both the phases of ATPS are largely water-based (80–85%), ATPS

http://dx.doi.org/10.1016/j.jchromb.2016.10.037
1570-0232/© 2016 Elsevier B.V. All rights reserved.
 A.M. Alhelli et al. / J. Chromatogr. B 1039 (2016) 66–73
creates a specific environment that partitions and concentrates the
biomolecules selectively into one of the phases, while preserving
the innate structure of the biomolecules [13]. The partitioning and
decontamination of enzymes in ATPS is primarily dependent on the
physico-chemical characteristics of the biomolecules, such as size,
shape, surface charge, molecular weight, and bonding interaction
and hydrophobicity (hydrogen bonding, electrostatic and Vander
Waals) with phase forming components. It is very important that
the partition parameters of the biomolecules are optimised in order
to achieve better recovery and increased purity of the enzymes
[9,14]. Polymer-salts ATPS have been utilised for the partitioning
and recovery of various lipase which offers several advantages like
low viscosity, low cost, and express phase separation compared
to polymer-polymer systems [15]. Hence, the polymer-salt system
(PEG-phosphate) was selected. However, there are some obsta-
cles in the use of ATPS due to its unpredictable functioning and
absence of data about the detailed mechanism of partitioning [16].
In recent times, the application of RSM in biological processes has
gained significant importance, which is appropriate and effective
for studying systems that consist of numerous independent vari-
ables with the potential to influence the final responses [17]. The
most important benefit of using RSM is that there is a considerable
decrease in the number of trials required to evaluate the impact of
the numerous variables on the interactions and responses [18]. In
this study, the prerequisites of lipase partition behaviour on differ-
ent variables, such as concentration of polymer, molecular weight
of polymer, concentration of salt and concentration of buffer were
taken into consideration. No recent information is available about
the ATPS extraction of the extracellular P. candidum (PCA 1/TT031)
lipase. Therefore, the objective of this study is to use RSM to analyse
the partitioning behaviour of lipase produced by P. candidum (PCA
1/TT031) in ATPS where PEG, hydrophilic polymer polyethylene
glycol (PEG), phosphate and NaCl interact to increase the purity of
the lipase.
2. Materials and methods
2.1. Feedstock preparation
The P. candidum (PCA 1/TT031) [commercially freeze–dried
strains from Chr. Hansen Sa (Arpajon, France)] was cultivated in
a 10 g wheat bran culture medium with the constituents: Tribu-
tyrin 2% (v/w), meat peptone 2% (w/w), moisture ratio 50% and
inoculum size 5 × 106 spore/gm wheat bran. The substrate’s ini-
tial pH value was attuned to 9 and the medium was incubated
at 20 ◦ C. After an incubation period of 7 days, the fermented cul-
ture was harvested with 50 ml of phosphate buffer pH 7. Later, the
mouldy bran was agitated using rotary shaker (180 rpm, 30 min) at
room temperature. To segregate the tiny elements of the substrate,
the broth extracts were accumulated and centrifuged (Refriger-
ated centrifuge SIGMA 3–18 K Goettingen, Germany) at 5000 × g
for 20 min at 4 ◦ C [19]. Specific activity of lipase in crude extract
was 5.9 U/mg protein. The clear and brown residual supernatant
was then used directly in the ATPS trials.
2.2. Lipase assay using oil emulsion method
The method suggested by Macedo, Park and Pastore [20] was
pursued that included measuring of the lipase activity with olive
oil as substrate. The substrate was prepared by preparing a fresh
concoction of oil emulsion solution [25 ml of olive oil with 75 ml
of Arabic gum solution (7%, w/v)]. While continuously stirring the
mixture till both the solutions were completely dissolved, the sub-
strate solution was formulated. In a conical flask a mixture of 4 ml of
100 mM sodium phosphate buffer (pH 8.0), 5 ml of the oil emulsion,
67
1 ml of 110 mM CaCl2 and 1 ml of enzyme extract was incubated
at 37 ◦ C for 30 min under shaking at 160 rpm. For every experi-
ment, proper blank was made where the enzyme was substituted
with distilled water. Under the assay environments, the reaction
was terminated by the addition of 15 ml ethanol: acetone mix-
ture (1:1 v/v).The liberated fatty acids were titrated with 50 mM
sodium hydroxide solution in the existence of phenolphthalein as
an indicator.
One unit (U) of enzyme activity was expressed as the amount of
1 ␮mol of free fatty acid released per minute per ml extract.
The Bradford method [21] was used to measure the concentra-
tion of protein in the samples and Bovine Serum Albumin (BSA)was
deemed as a standard.
2.3. ATPS preparation
–
Peričin, Madarev-Popovi ć and Vaštag [22] suggested that ATPS
was formulated with a few alterations in 15 ml graduated tubes that
have a final mass of 10 g attained by mixing a proportionate amount
of a stock solution of PEG 50% (w/w), sodium chloride 20% (w/w),
and potassium phosphate (mixture of KH2 PO4 and K2 HPO4 ) 40%
(w/w) with pH 7. There was an addition of crude extract to the final
concentration of 20% (w/w). To attain a final weight of 10 g at room
temperature 25 ◦ C, distilled water was also added. To accelerate the
phase separation, the content was carefully mixed with the help of
a vortex mixer for 10 min and then centrifuged at 10,000 rpm for
10 min at 25 ◦ C [23]. As a result of this, the two phases became
clear and evident, and the interface became absolutely distinct.
The phases were then methodically removed with a long-needle
syringe for the bottom phase, and with a pipette for the top phase.
The volume of each phase was measured, and samples from the two
phases were precisely examined and tested for total protein con-
centration and enzyme assay [24]. To avoid interaction among the
phase components, samples were analysed against blanks without
enzymes containing the same components.
2.4. Determination of partition coefficient, purification factor and
yield
The partition coefficient (Ke ) of lipase is defined as the ratio of
the lipase activity in the two phases and is represented with the
following equation (Eq. (1)):
Ke = AT /AB (1)
The ratio of the observed specific activity of lipase in the top
phase to the initial specific activity of lipase in the crude extract
is referred to as the top phase purification-fold (PFT ), and is repre-
sented with (Eq. (2)):
PFT = Specific activity of top phase sample
/Specific activity of crude feedstock (2)
The ratio of lipase activity to the total protein concentration of
the sample denotes the specific activity. Since enzyme was pref-
erentially partitioned to the top phase, the yield of lipase was
calculated for this phase with the help of Eq. (3):
Y T (%) = AT /AI × 100 (3)
where AT and AI are the lipase activities in top phase and initial
enzyme activity, respectively [14].
2.5. Experimental design
The objective of the experiment is to determine the impact
of four independent variables of the ATPS approach using RSM,
68 A.M. Alhelli et al. / J. Chromatogr. B 1039 (2016) 66–73

Table 1
Experimental design from the central composite design (CCD).

Run order Blocks Independent variables

Molecular mass Concentration of Concentration of Concentration of

of PEG(g/mol, X1) PEG% (w/w, X2) buffer% (w/w, NaCl% (w/w, X4)
X3)

1 3 6000 10.5 12.0 5.2

2 3 6000 10.5 15.0 2.6
3 3 6000 10.5 12.0 0.0
4 3 6000 9 12.0 2.6
5 3 6000 10.5 9.0 2.6
6 3 10,000 10.5 12.0 2.6
7 3 1500 10.5 12.0 2.6
8 3 6000 12 12.0 2.6
9c 3 6000 10.5 12.0 2.6
10c 3 6000 10.5 12.0 2.6
11 1 4000 9.7 13.5 1.3
12 1 4000 11.2 10.5 1.3
13 1 4000 11.2 13.5 3.9
14c 1 6000 10.5 12.0 2.6
15 1 8000 9.7 13.5 3.9
16 1 4000 9.7 105 3.9
17c 1 6000 10.5 12.0 2.6
18 1 8000 11.2 10.5 3.9
19 1 8000 11.2 13.5 1.3
20 1 8000 9.7 10.5 1.3
21 2 4000 9.7 10.5 1.3
22 2 8000 11.2 10.5 1.3
23 2 8000 11.2 13.5 3.9
24c 2 6000 10.5 12.0 2.6
25 2 4000 11.2 10.5 3.9
26 2 8000 9.7 13.5 3.9
27 2 4000 9.7 13.5 3.9
28c 2 6000 10.5 12.0 2.6
29 2 8000 9.7 13.5 1.3
30 2 4000 11.2 13.5 1.3
c
Center point.

which includes concentration of phosphate buffer (9–15% (w/w),
X3), concentration of PEG (9–12% (w/w), X2), molecular mass of
PEG (1500–10,000 (g/mol), X1), and concentration of NaCl (0–5.2%
(w/w), X4) on purification factor (Y2), partition coefficient (Y1), and
yield (Y3) of the purified lipase obtained from P. candidum (PCA
1/TT031). The central composite design (CCD) was used to evalu-
ate and estimate the full quadratic model for each response. On the
basis of CCD, thirty supernatants were assessed with four indepen-
dent variables at five different levels of each variable (Table 1). To
evaluate the data statistically and graphically, the statistical soft-
ware Minitab v16 statistical package (Minitab Inc., PA, USA) was
used.
2.6. Variance analysis
To determine the significant variables in addition to LSD or to
assess the least significance tests to measure the variance amongst
the studied samples, the Analysis of Variance (ANOVA) technique
was used. Every sample was subjected to two fold or three fold mea-
surement and the subsequent data was recorded and documented
as the mean ±SD of independent tests. Fisher’s test (F-test) analysed
the consequence of the suggested model. Moreover, to calculate
the coefficient of determination (R2 ), the regression equations were
summated to the F-test. To envisage the relationships between the
dependent variables (responses) and the experimental levels of the
independent variables (different factors) involved in the design,
the corresponding polynomial equation was expressed as three-
dimensional surface plots. To endorse the validity and suitability of
the model, different permutations (graphical and numerical) of the
optimised parameters were tested experimentally [25].
3. Results and discussion
3.1. Fitting the response surface models
Table 2 demonstrates the expected regression coefficients of the
response surface models and the corresponding R2 values. As indi-
cated in Table 2, the significant (p ≤ 0.05) response surface models
were achieved for all the response variables with a high R2 value
in the diverse range of 0.90–0.94. Accordingly, it can be concluded
that the models of response surface were appropriately and accu-
rately used for estimating the high variation percentage (<80%) of
lipase properties in relation to the purification variables. Table 3
demonstrates the main, quadratic and interaction consequences of
the PEG molecular mass (X1), concentration of phosphate buffer
(X3), concentration of PEG (X2) and the concentration of NaCl (X4)
on every response variable. Also, Table 3 shows the importance
of the p-value and F-ratio. The results shown in Table 3 indicate
that the independent variables had the most significant (p ≤ 0.05)
impact on the purification factor, partition coefficient, and the yield
of the purified lipase from P. candidum (PCA 1/TT031).
3.2. Partition coefficient (Y1)
The extraction efficiency is affected by the PEGs due to their
different degrees of polymerisation. They alter the phase dia-
gram by changing the constituents of phases and the number of
polymer–enzyme interactions [26]. Table 3 indicates that the par-
tition coefficient (Y1) was considerably (p ≤ 0.05) affected by the
main effect of PEG molecular mass, PEG concentration, buffer con-
centration and NaCl concentration; the quadratic effects of PEG
concentration, buffer and NaCl concentration; and the interactions
 A.M. Alhelli et al. / J. Chromatogr. B 1039 (2016) 66–73 69

Table 2
Regression coefficient, R2 , probability values of the response surface models.

Regression coefficient Partition coefficient (Y1) Purification factor (Y2) Yield (%,Y3)

b0 136.443 425.214 1370.84

b1 0.003 −0.011 0.04
b2 −19.069 −74.408 −171.69
b3 −5.367 4.09 −68.62
b4 −6.347 −5.976 −61.28
b1 2 0.000 0.000 0.000
b2 2 0.823 4.186 8.83
b3 2 0.13 0.434 2.38
b4 2 0.279 0.151 3.16
b12 −0.000 0.000 −0.000
b13 −0.000 −0.000 −0.000
b14 0.000 0.000 0.000
b23 0.211 −1.378 0.72
b24 0.149 −0.242 −0.96
b34 0.22 0.545 3.81
R2 0.90 0.94 0.92
Regression P- value 0.000* 0.000* 0.000*

bi: the estimated regression coefficient for the main linear effects. bii: the estimated regression coefficient for the quadratic effects. bij: the estimated regression coefficient
for the interaction effects. 1: molecular mass of PEG; 2: concentration of PEG; 3: concentration of buffer; 4: concentration of NaCl.
*
Significant (p ≤ 0.05).

Fig. 1. Response surface plots for interaction effects of ATPS partition coefficient on enzymatic properties of lipase.

of PEG molecular mass and PEG concentration, buffer and NaCl
concentration in the ATPS. The results revealed that the most sig-
nificant (p ≤ 0.05) effect on partition coefficient value (Y1) of the
alkaline lipase was caused by the primary and quadratic impact
of concentration of PEG and salt, respectively. Figs. 1 and 2 show
the response surface plots in optimizing the partition coefficient of
lipase enzyme in the top PEG phase.
It is evident from the results presented in Fig. 1 that the inter-
actions between the PEG molecules-3000 (g/mol) and 9.2% (w/w)
concentration have led to an increase in the partition coefficient
value (Ke = 4.0). This can be probably due to the reason that at this
molecular weight (MW), there is an increase in preferential inter-
actions between the PEG and the protein area, which results in
an increase in the partition coefficient of the top phase. The high
molecular mass PEG is selectively partitioned to the bottom phase
to achieve the steric prohibition of the hydrophilic protein from
the bottom phase to the top phase. The salting-out forces that drive
most proteins to the salt-rich bottom phase reject the volume effect
of the PEG chain that progressively controls the volume of the top
phase. This allows very less space for the protein in the bottom
phase [9,27,28]. Besides, the interfacial tension between the phases
may be affected by the variance partitioning of the biomolecules
[29]. In addition, the interfacial tension can be decreased by reduc-
ing the molecular weight of the phase-forming polymers [30,31]. As

Fig. 2. Response surface plots for interaction effects of ATPS partition coefficient on enzymatic properties of lipase.
70 A.M. Alhelli et al. / J. Chromatogr. B 1039 (2016) 66–73

X1, X2,X3 and X4: The main effect of PEG molecular mass, PEG concentration, buffer concentration and NaCl, respectively X12 , X22 , X32 and X42 : The quadratic effect of PEG molecular mass, PEG concentration, buffer concentration
and NaCl, respectively. X1X2: The interaction effect of PEG molecular mass and PEG concentration, X1X3: The interaction effect of PEG molecular mass and buffer concentration, X1X4: The interaction effect of PEG molecular
principle, the partitioning of enzymes in the top phase decreases as

0.012*

0.003*
13.49
X3X4
the PEG molecular mass increases. Hence, the selection of appropri-

8.63

2.76
0.12
ate molecular mass of polymers in this system is highly significant.
In addition, Fig. 2 shows that the partition coefficient was affected
(p ≤ 0.05) by the interaction of buffer and NaCl concentration. ATPS

0.337

0.718

0.651
X2X4

0.99

0.14

0.21
with the 13.8% (w/w) phosphate buffer led to higher partition coef-
ficient. Bradoo, Saxena and Gupta [32], Bandmann et al., [33] and
Gulati, Saxena and Gupta [34] observed that in ATPS, the phosphate

mass and NaCl, X2X3: The interaction effect of PEG concentration and buffer, X2X4: The interaction effect of PEG concentration and NaCl, X3X4: The interaction effect of buffer concentration and NaCl.
system for lipase separation, concentration and purification is more
X2X 3

0.129

0.695
0.03*
2.63

5.89

0.16
appropriate due to its considerable influence on the system envi-
ronment. Furthermore, as salts are strongly bonded with a large
amount of water molecules, the addition of salt in high concentra-
0.218

0.269

0.139
X1X4

tions results in protein aggregation and precipitation. As a result,

1.68

1.33

2.48

the interactions between proteins and salts become stronger than

the interactions between protein and water, which imply dehydra-
tion of proteins [35,36].
0.461

0.188
0.091
X1X3
Interaction effects

3.33

1.93
0.58

According to Huddleston et al. [37] and Barbosa et al. [38],

there is a fine balance between the PEG hydrophobicity and the
salting-out ability of salts. The influence of the PEG molecular
mass on the composition of phases and the number of poly-
0.024*

0.005*
11.74
0.385
X1X2

6.51

0.81

mer protein interactions has been revealed in various studies

[39].
0.001*

0.004*
17.85

11.94
0.611
0.27
2

3.3. Purification factors (Y2)

X4

When employing ATPS to purify lipases, the primary effects of

0.021*

0.004*

PEG molecular masses and concentrations, the quadratic results of

11.94
0.067
6.83

4.00
2
X3

PEG molecular masses and PEG concentrations, as well as the inter-

actions of PEG concentrations and buffer concentrations, all had
a considerable (p ≤ 0.05) impact on factor values for purification
0.001*

0.007*
0.000*

(Table 3).
17.19

23.29

10.32
Quadratic effects
2
X2

The quadratic influence of PEG molecular masses and concen-

trations was of the greatest significance (p ≤ 0.05) in purification
factors for lipases (Y2). PEG molecular weights have been con-
0.027*
0.001*
21.05
0.909

firmed to influence PEG distributions in both phases and in

6.18
0.01
2
X1

polymer/protein exchanges [39]. For a PEG with a molecular weight

of 3000 (g/mol), purification factor was observed to be maxi-
mum.
0.005*

0.034*
11.49
0.478

The maximum purification factor of 33.9 was obtained by using

5.58
0.53
X4

9.2% (w/w) PEG, 13.8% (w/w) buffer concentration. Fig. 3 illustrates

the process by which interactions among independent variables
(PEG and buffer concentration) extensively affect (p ≤ 0.05) the
0.012*

0.019*

purification factors of lipases derived from P. candidum (PCA

8.48

7.23
0.62
0.26
X3

1/TT031). This could be due to an increase in polymer concentra-

tions with regards to viscosity and high density of the phase. So, a
high concentration of polymer offers large variance in properties
0.002*

0.016*
0.001*
17.94

14.31

7.58

between the phases of the ATPS [9]. As such, to transfer the lipase
X2

molecules to the top PEG phase the purification factor is raised.

Main effects
Analysis of variance for response surface quadratic model.

Purification factor for lipase is decreased with increased phosphate

and PEG concentration, because of more rapid total protein parti-
0.028*

0.024*

0.04*
6.12

5.08

tioning in the top phase instead of lipase.

5.2
X1

The maximum purification factor was attained at 3.3% NaCl

concentration. In general, adding salt in ATPS affects the partition-
ing of proteins in terms of differences in electrostatic potentials.
p-value

p-value

p-value
F-ratio

F-ratio

F-ratio

This can be attributed to the variation in ion affinities in both the

phases [40]. Ion partiality for either hydrophobic micellar phase or
hydrophilic aqueous phase is determined from their hydrophobic
Significant at (p < 0.05).

properties. Proteins with higher hydrophobic levels tend to parti-

Purification factor (Y2)

tion to micellar phases and vice-versa [40]. The addition of NaCl

Partition coefficient

generally improves the hydrophilicity of phases rich in PEG [41], or

interphases, while the greater intensity of hydrophobic exchanges
Yield (Y3, %)

between protein and PEG molecules apparently provides a possible

Variables

explanation for the improvement in factors for purification (salting-

Table 3

(Y1)

out process) [42]. This implies that there is an increase in disparities

*

in hydrophobicity between each phase.

 A.M. Alhelli et al. / J. Chromatogr. B 1039 (2016) 66–73 71

Fig. 3. Response surface plots for interaction effects of ATPS purification factor on enzymatic properties of lipase.

Fig. 4. Response surface plots for interaction effects of ATPS yield on enzymatic properties of lipase.

Fig. 5. Response surface plots for interaction effects of ATPS yield on enzymatic properties of lipase.

3.4. Yields (Y3)
Lipase yields were considerably influenced (p ≤ 0.05) by ATPS
preconditions such as PEG molecular masses, PEG concentration,
buffer, NaCl concentrations and by the quadratic impacts of PEG
molecular masses, PEG concentrations, buffer concentrations and
NaCl concentrations, as well as by the interactions between PEG
molecular masses and concentrations, buffer concentration and
NaCl concentration in the ATPS. Table 3 illustrates the interac-
tion impact of phosphate concentration and NaCl concentration,
which exhibited the greatest influence (p ≤ 0.05) on yields (Y3). An
increase in PEG molecular masses leads to a subsequent increase
in polymer chain lengths, which decreases the free volume [43].
Thus, biomolecules will partition to the bottom phase. The analysis
of the purification factor’s curve and shape (Figs. 4 and 5) shows
that variations in yields (Y3) result from the non-linear function-
ing of purification preconditions. There was a significant (p ≤ 0.05)
increase in the yield of lipase with the addition of 9.2% (w/w) PEG
concentration. Also, with increasing PEG concentration, the yield
improvement effect tended to increase. Given the increase in the
molecular masses of PEG, yields also showed an improvement. A
rise in PEG molecular masses enhanced the effects of PEG on lipase,
which led to a subsequent increase in PEG hydrophobic qualities
[44]. These effects are attributed to the salting-out processes of
potassium phosphate buffers that results in the molecular elimina-
tion of lipase, which moves to the upper phase. Parametric values
show that proteins and enzymes partitioned (K > 1) into the upper
phase, which manifested a greater hydrophilicity as a result of the
presence of dominant PEG. This indicates that the enzymes have a
greater partiality for less polar environments [45].
Considerable yields of lipases were observed when salts were
present at 3.3% (w/w). In contrast, yield decreased at higher concen-
72 A.M. Alhelli et al. / J. Chromatogr. B 1039 (2016) 66–73

Fig. 6. Fitted line plot for predicted (Y1) and experimental values (Y0) of Partition Fig. 7. Fitted line plot for predicted (Y1) and experimental values (Y0) of purification
coefficient. factor.

trations of NaCl (>3.3%). This could be justified by the fact that the
crude extract impurities partitioned together with lipase toward
the top. The addition of salts in ATPS processes can therefore be
employed for the efficient separation of lipase.
Andrews, Schmidt and Asenjo [46] described how adding salts
can result in protein transfers from one phase to another as a func-
tion of the hydrophobic disparities between ATPS phases, charges
on the protein surfaces, as well as hydrophobicity.

3.5. Optimisation procedures

Multiple-response optimisations were performed to ascer-

tain optimal levels for each independent variable to achieve
desired responses [47]. As reported by Montgomery [48], three-
dimensional, or 3D, response surface plotting can be used to explain
interactions between fixed and dependent variables.
Optimal preconditions were obtained by using numerical opti-
misations to precisely assess the optimal values of each response
variable. The results showed that purification preconditions with
PEG combinations of 3000 (g/mol) molecular mass, 9.2% (w/w)
concentration, 13.8% (w/w) buffer concentration and NaCl 3.3%
(w/w) concentration were optimal for all lipase properties. The
corresponding response variables for the partition coefficients,
purification factors, and lipase yields were estimated to be 4.0, 33.9
and 84.0% under these optimal preconditions.
3.6. Experimental confirmation of models
Response surface equations were sufficiently validated by the
contrast between experimental values and the estimated values
derived from the response regression [49]. These outcomes were
mirrored by the general agreement among the variables used in
estimating variances in the properties of enzymes, which occur
due to purification preconditions. Thus, the nearer the experimen-
tal values to the estimated values, the better the explanation for
the fitness of the regression [49]. In Figs. 6–8 , it is shown that the
respective values of the response variables obtained from observa-
tions were nearer to those estimated in the modelled equations.
Fig. 8. Fitted line plot for predicted (Y1) and experimental values (Y0) of yield%.
aqueous two-phase methodology for the extraction and purifica-
tion of lipase was PEG-phosphates aqueous two-phase systems
that support the top phase, which is constituted of 9.2% (w/w) PEG
concentration, 3000 (g/mol) molecular mass, 3.3% (w/w) NaCl con-
centration and 13.8% (w/w) phosphate buffer concentration. Under
optimal prerequisite conditions, the corresponding response vari-
ables for the purification factor, partition coefficient, and lipase
yield were assessed to be 33.9, 4.0 and 84.0%. For optimising lipase
purification in PEG-phosphate-NaCl ATPS, the use of statistical
experimental design proved to be beneficial. The purification of
enzyme is greatly influenced by the phase-forming polymer’s con-
centration and molecular mass. Moreover, an agreement between
the experimental and projected results was attained, which vali-
dated the proposed model. In this paper, the collective viability of
ATPS optimisation with the use of RSM for recovery and purification
of lipase is shown.

4. Conclusion Acknowledgments

For the first time, the aqueous two-phase extraction was used We thank the ministry of higher education Malaysia for finan-
for simultaneous partitioning and purification of alkaline lipase cial support this project and the ministry of higher education and
from P. candidum (PCA 1/TT031). It was found that the most suitable scientific research of Iraq for financial support to Amaal M. Alhelli.
 A.M. Alhelli et al. / J. Chromatogr. B 1039 (2016) 66–73
Special thanks are extended to Faculty of Food Science and Tech-
nology, University Putra Malaysia.
References
[1] E. Dalmau, J. Montesinos, M. Lotti, C. Casas, Effect of different carbon sources
on lipase production by Candida rugosa, Enzyme Microb. Technol. 26 (2000)
657–663.
[2] X.-G. Gaoa, S.-G. Cao, K.-C. Zhang, Production, properties and application to
nonaqueous enzymatic catalysis of lipase from a newly isolated Pseudomonas
strain, Enzyme Microb. Technol. 27 (2000) 74–82.
[3] M.A.F. Costa, R.M. Peralta, Production of lipase by soil fungi and partial
characterization of lipase from a selected strain (Penicillium wortmanii), J.
Basic Microbiol. 39 (1999) 11–15.
[4] N. Gupta, V. Sahai, R. Gupta, Alkaline lipase from a novel strain Burkholderia
multivorans: statistical medium optimization and production in a bioreactor,
Process Biochem. 42 (2007) 518–526.
[5] L.P.G. Franken, N.S. Marcon, H. Treichel, D. Oliveira, D.M. Freire, C. Dariva, J.
Destain, J.V. Oliveira, Effect of treatment with compressed propane on lipases
hydrolytic activity, Food Bioprocess Technol. 3 (2010) 511–520.
[6] J.N.d. Prazeres, J.A.B. Cruz, G.M. Pastore, Characterization of alkaline lipase
from Fusarium oxysporum and the effect of different surfactants and
detergents on the enzyme activity, Braz. J. Microbiol. 37 (2006) 505–509.
[7] B. Ruiz, A. Farrés, E. Langley, F. Masso, S. Sánchez, Purification and
characterization of an extracellular lipase from Penicillium candidum, Lipids
36 (2001) 283–289.
[8] H.S. Oberoi, Y. Chavan, S. Bansal, G.S. Dhillon, Production of cellulases through
solid state fermentation using kinnow pulp as a major substrate, Food
Bioprocess Technol. 3 (2010) 528–536.
[9] K. Ratanapongleka, Recovery of biological products in aqueous two phase
systems, Int. J. Chem. Eng. Appl. 1 (2010) 191.
[10] M. Li, J.-W. Kim, T.L. Peeples, Amylase partitioning and extractive
bioconversion of starch using thermoseparating aqueous two-phase systems,
J. Biotechnol. 93 (2002) 15–26.
[11] B.Y. Zaslavsky, Aqueous Two-phase Partitioning: Physical Chemistry and
Bioanalytical Applications, CRC Press, 1994.
[12] G. Tubio, G.A. Picó, B.B. Nerli, Extraction of trypsin from bovine pancreas by
applying polyethyleneglycol/sodium citrate aqueous two-phase systems, J.
Chromatogr. B 877 (2009) 115–120.
[13] C.W. Ooi, B.T. Tey, S.L. Hii, A. Ariff, H.S. Wu, J.C.W. Lan, R.S. Juang, S.M.M.
Kamal, T.C. Ling, Direct purification of Burkholderia pseudomallei lipase from
fermentation broth using aqueous two-phase systems, Biotechnol. Bioprocess
Eng. 14 (2009) 811–818.
[14] P.K. Srivastava, M. Kapoor, Extracellular endo-mannanase from Bacillus sp.
CFR1601: economical production using response surface methodology and
downstream processing using aqueous two phase system, Food Bioprod.
Process. 91 (2013) 672–681.
[15] M.T.H. Cavalcanti, T.S. Porto, B. de Barros Neto, J.L. Lima-Filho, A.L.F. Porto, A.
Pessoa, Aqueous two-phase systems extraction of ␣-toxin from Clostridium
perfringens type A, J. Chromatogr. B 833 (2006) 135–140.
[16] H.S. Mohamadi, E. Omidinia, R. Dinarvand, Evaluation of recombinant
phenylalanine dehydrogenase behavior in aqueous two-phase partitioning,
Process Biochem. 42 (2007) 1296–1301.
[17] J. Singh, A. Sharma, Application of response surface methodology to the
modeling of cellulase purification by solvent extraction, Sci. Res. 3 (2012).
[18] P. Singh, S.S. Shera, J. Banik, R.M. Banik, Optimization of cultural conditions
using response surface methodology versus artificial neural network and
modeling of l-glutaminase production by Bacillus cereus MTCC 1305,
Bioresour. Technol. 137 (2013) 261–269.
[19] N.D. Mahadik, U.S. Puntambekar, K.B. Bastawde, J.M. Khire, D.V. Gokhale,
Production of acidic lipase by Aspergillus niger in solid state fermentation,
Process Biochem. 38 (2002) 715–721.
[20] G. Macedo, Y. Park, G. Pastore, Partial purification and characterization of an
extracellular lipase from a newly isolated strain of Geotrichum sp, Rev.
Microbiol. 28 (1997) 90–95.
[21] M.M. Bradford, A rapid and sensitive method for the quantitation of
microgram quantities of protein utilizing the principle of protein-dye binding,
Anal. Biochem. 72 (1976) 248–254.
–
[22] D.M. Peričin, S.Z. Madarev-Popovi ć, Ž.G. Vaštag, Partitioning of acid protease
from Penicillium roqueforti in aqueous two-phase system polyethylene
glicol/phosphate, Acta Periodica Technol. (2008) 171–180.
[23] C.W. Ooi, C.P. Tan, S.L. Hii, A. Ariff, S. Ibrahim, T.C. Ling, Primary recovery of
lipase derived from Burkholderia sp. ST8 with aqueous micellar two-phase
system, Process Biochem. 46 (2011) 1847–1852.
[24] A.-J. Tong, J.-J. Dong, L.-D. Li, Aqueous two-phase extraction system of sodium
perfluorooctanoate and dodecyltriethylammonium bromide mixture and its
application to porphyrins and dyes, Anal. Chim. Acta 390 (1999) 125–131.
73
[25] Q. Xu, Y. Shen, H. Wang, N. Zhang, S. Xu, L. Zhang, Application of response
surface methodology to optimise extraction of flavonoids from fructus
sophorae, Food Chem. 138 (2013) 2122–2129.
[26] P. Albertson, Partition of Cell Particles and Macromolecules, 3rd edn., Wiley &
Sons, New York, 1986.
[27] R. Hatti-Kaul, Aqueous Two-phase Systems: Methods and Protocols, Springer
Science & Business Media, 2000.
[28] L. Yang, D. Huo, C. Hou, K. He, F. Lv, H. Fa, X. Luo, Purification of plant-esterase
in PEG1000/NaH 2 PO 4 aqueous two-phase system by a two-step extraction,
Process Biochem. 45 (2010) 1664–1671.
[29] N. Srinivas, R. Barhate, K. Raghavarao, Aqueous two-phase extraction in
combination with ultrafiltration for downstream processing of Ipomoea
peroxidase, J. Food Eng. 54 (2002) 1–6.
[30] Y.-T. Wu, Z.-Q. Zhu, L.-H. Mei, Interfacial tension of poly (ethylene glycol)+
salt+ water systems, J. Chem. Eng. Data 41 (1996) 1032–1035.
[31] T. Karka&#x015F, S. Önal, Characteristics of invertase partitioned in poly
(ethylene glycol)/magnesium sulfate aqueous two-phase system, Biochem.
Eng. J. 60 (2012) 142–150.
[32] S. Bradoo, R. Saxena, R. Gupta, Partitioning and resolution of mixture of two
lipases from Bacillus stearothermophilus SB-1 in aqueous two-phase system,
Process Biochem. 35 (1999) 57–62.
[33] N. Bandmann, E. Collet, J. Leijen, M. Uhlén, A. Veide, P.-Å. Nygren, Genetic
engineering of the Fusarium solani pisi lipase cutinase for enhanced
partitioning in PEG-phosphate aqueous two-phase systems, J. Biotechnol. 79
(2000) 161–172.
[34] R. Gulati, R. Saxena, R. Gupta, Fermentation and downstream processing of
lipase from Aspergillus terreus, Process Biochem. 36 (2000) 149–155.
[35] R.C. Bonomo, L.A. Minim, J.S. Coimbra, R.C. Fontan, L.H.M. da Silva, V.P. Minim,
Hydrophobic interaction adsorption of whey proteins: effect of temperature
and salt concentration and thermodynamic analysis, J. Chromatogr. B 844
(2006) 6–14.
[36] H.S. Mohammadi, S.S. Mostafavi, S. Soleimani, S. Bozorgian, M. Pooraskari, A.
Kianmehr, Response surface methodology to optimize partition and
purification of two recombinant oxidoreductase enzymes, glucose
dehydrogenase and d-galactose dehydrogenase in aqueous two-phase
systems, Protein Exp. Purif. 108 (2015) 41–47.
[37] J.G. Huddleston, K.W. Ottomar, D.M. Ngonyani, A. Lyddiatt, Influence of
system and molecular parameters upon fractionation of intracellular proteins
from Saccharomyces by aqueous two-phase partition, Enzyme Microb.
Technol. 13 (1991) 24–32.
[38] J.M.P. Barbosa, R.L. Souza, A.T. Fricks, G.M. Zanin, C.M.F. Soares, Á.S. Lima,
Purification of lipase produced by a new source of Bacillus in submerged
fermentation using an aqueous two-phase system, J. Chromatogr. B 879
(2011) 3853–3858.
[39] D. Forciniti, C.K. Hall, M. Kula, Protein partitioning at the isoelectric point:
influence of polymer molecular weight and concentration and protein size,
Biotechnol. Bioeng. 38 (1991) 986–994.
[40] H.O. Johansson, J. Persson, F. Tjerneld, Thermoseparating water/polymer
system: a novel one-polymer aqueous two-phase system for protein
purification, Biotechnol. Bioeng. 66 (1999) 247–257.
[41] A. Hemavathi, K. Raghavarao, Differential partitioning of ␤-galactosidase and
␤-glucosidase using aqueous two phase extraction, Process Biochem. 46
(2011) 649–655.
[42] T. Franco, A. Andrews, J. Asenjo, Use of chemically modified proteins to study
the effect of a single protein property on partitioning in aqueous two-phase
systems: effect of surface hydrophobicity, Biotechnol. Bioeng. 49 (1996)
300–308.
[43] G. Patil, K. Raghavarao, Aqueous two phase extraction for purification of
C-phycocyanin, Biochem. Eng. J. 34 (2007) 156–164.
[44] G. Picó, D. Romanini, B. Nerli, B. Farruggia, Polyethyleneglycol molecular mass
and polydispersivity effect on protein partitioning in aqueous two-phase
systems, J. Chromatogr. B 830 (2006) 286–292.
[45] J.A. Asenjo, B.A. Andrews, Aqueous two-phase systems for protein separation:
a perspective, J. Chromatogr. A 1218 (2011) 8826–8835.
[46] B. Andrews, A. Schmidt, J. Asenjo, Correlation for the partition behavior of
proteins in aqueous two-phase systems: effect of surface hydrophobicity and
charge, Biotechnol. Bioeng. 90 (2005) 380–390.
[47] H. Mirhosseini, C.P. Tan, Response surface methodology and multivariate
analysis of equilibrium headspace concentration of orange beverage emulsion
as function of emulsion composition and structure, Food Chem. 115 (2009)
324–333.
[48] D. Montgomery, Design and Analysis of Experiments, 5th edn., Wiley, New
York USA, 2001.
[49] H. Mirhosseini, C.P. Tan, N.S. Hamid, S. Yusof, Effect of Arabic gum, xanthan
gum and orange oil on flavor release from diluted orange beverage emulsion,
Food Chem. 107 (2008) 1161–1172.